Benchmarking of mobile apps on heart failure Benchmarking de aplicativos móveis sobre insuficiência cardíaca Benchmarking de aplicaciones móviles sobre insuficiencia cardíaca REVIEW

Rev Bras Enferm. 2022;75(1): e20201093 https://doi.org/10.1590/0034-7167-2020-1093 10 of ABSTRACT Objective: to analyze the mobile apps on heart failure available in the main operating systems and their usability. Methods: benchmarking of mobile applications, systematic research, comprising 38 mobile applications for analysis of general information, functionalities and usability. Usability was assessed using System Usability Scale and Smartphone Usability Questionnaire, followed by the calculation of the agreement index and the exact binomial distribution test, with a significance level of p> 0.05 and a proportion of 0.90. Results: mobile applications had English as the predominant language (73.7%), were directed to patients (71.1%) and the predominant theme was disease knowledge (34.2%). Functionalities ranged from general features to the need for an internet connection. In assessing usability, heart failure was shown to be 92.1% -94.7% and p <0.05. Final considerations: the mobile apps on heart failure have varied content and adequate usability. However, there is a need to develop more comprehensive mobile applications. Descriptors: Heart Failure; Mobile Applications; Telemedicine; Health Promotion; Diffusion of Innovation..


INTRODUCTION
Heart failure (HF) is an emerging global threat, with a current prevalence of 64.34 million cases on the planet (8.52 per 1,000 inhabitants), representing 9.91 million years lost due to disability and spending of US$346.17 billion (1) , with a prospect of an increase despite therapeutic advances. These data alert to the prioritization of preventive actions and {the imminence of considering new ways of care.
In order to facilitate the process of caring for people with HF, mHealth technologies, such as mobile applications (APP), stand out. In recent years, the number of APPs has increased exponentially in the most varied areas and has contributed to increase productivity and quality of health care, in addition to allowing the use of the most up-to-date clinical knowledge and supporting the clinical decision-making of professionals (2)(3) . When undertaken in the care of patients with HF, APPs contribute to survival and improved quality of life (4)(5)(6) .
After creating an APP, it is necessary to evaluate its content and functionality, to ensure the launch of a product suitable for the target audience. The development demands an understanding of functionalities and previous knowledge of APP already built for the theme. In this process, benchmarking is a technique that allows assessment of performance and results of this technology, analyzing, conclusively, whether they are positive or negative (7) .
The evaluative concern was observed in studies that reviewed self-care APP for patients with HF (8)(9) , which corroborate that the existing ones are incomplete, of low quality, with bugs and unsuitable to the target audience, having little impact in disease management, in addition to excluding professionals and family members/caregivers. Still, they address only self-care, an important element in the complexity of a person with HF (10) , but not the only one.
The need to expand the review of APPs is highlighted, to cover all content and resources produced about HF and to involve all participants in the care process. Thus, it was proposed a systematic survey of all APPs about HF currently available in the virtual stores Play Store and App Store, aimed at patients and their families/caregivers and health professionals. Investigating its functionality and usability is essential to maximize its potential health promoter.

OBJECTIVE
To analyze the mobile apps on heart failure available in the main operating systems and their usability.

Ethical aspects
This study did not require prior approval from an Institutional Review Board. Even so, the researchers' ethical commitment is reiterated, as recommended by Resolution 466/12 of the Brazilian National Research Council (Conselho Nacional de Saúde).

Type of study
This is a benchmarking of mobile applications, a systematic survey that sought to identify all the APPs about HF and their usability in the main operating systems.

Methodological procedures
In order to systematize the data collection for conducting benchmarking, the steps of the Preferred Reporting Items for Systematic Reviews and Meta-Analyzes (PRISMA) checklist were followed. The steps taken were: 1) establishment of analytical objectives; 2) choice of terms; 3) enumeration of APP inclusion/ exclusion criteria; 4) definition of information to be extracted; 5) analysis of results; 6) presentation of results and discussion.
Initially, the objectives of analysis were established: to identify APP over HF, download, analyze content and functionalities and classify them according to usability. The search was made easier with the use of heart failure terms.
APPs were included in the survey in any language that covered the HF theme and excluded if 1) required payment for installation, 2) needed author release for use (registration number and password), 3) were intended for event monitoring/scientific conference, 4) contained only books or guidelines, 5) acted as general health calculators, unspecific for HF and 6) provided only vital signs.

Data source
Searches for APPs were carried out in the Play Store and App Store virtual stores of iOS (Apple) and Android (Google) operating systems, respectively.

Data collection and organization
Two independent reviewers performed the screening of APPs in virtual stores. For this, three devices were used: a Samsung Galaxy S8, compatible with Android; an iPhone XS Max and an iPad Pro, version iPadOS 13.5.1, compatible with iOS.
Four searches were carried out in each virtual store, using each of the previously established terms individually. After the research, the APPs were previously selected via title and short description. This screening served to separate most APPs on cardiac health, in a generalized way. APPs that met the inclusion and exclusion criteria were downloaded and analyzed.
Each reviewer organized the APPs in an information matrix in Microsoft Excel ® , with description of identification (name, developer, author (s) responsible, year of launch, language, country of origin), target audience (patient, family/caregivers and students or health professionals), APP data (description, current version, number of downloads, rating/evaluation), content, features (features) and usability.
Draft of the data extraction tool was modified and revised, as needed, during the data extraction process for each included APP. Additional information and comments were recorded individually for further analysis.
APPs were tested for two weeks, as some needed control for a minimum period of one week to provide data (graphs) to users. At the end of that period, the researchers felt safe to collect the data and assess its usability.

Data analysis
The two independent reviewers carried out usability assessment mediated by two instruments. The first was the System Usability Scale (SUS) questionnaire, created by John Brooke, in 1986, and validated in Brazil by Tenório (11) . The SUS questionnaire is used to evaluate products, services, hardware, software, websites and applications. It consists of ten questions and, for each one, users can answer on a scale of 1 (I strongly disagree) to 5 (I strongly agree).
For usability calculation, 1 is subtracted from the score for odd answers, and for even answers, the answer is subtracted from 5. To obtain the final average, multiply the value found by 2.5, which will give the final score (between 0 to 100). The SUS average is 68 points; below this value, the product has usability problems. Software that scores above 85 has excellent acceptance by users (12) .
The second questionnaire was version 1.0 of Smartphone Usability questionnaiRE (SURE), built and validated by Wangenheim et al. (11) after an exhaustive systematic literature review and the use of Item Response Theory (IRT) for the construction of its items. This questionnaire has 31 items and measures the usability of a smartphone APP at levels ranging from 30 (all respondents are more likely to disagree partially or totally with the items) to 80 (respondents begin to fully agree that the help/tip given by APP was helpful) (13) .
The variables contained in the information matrix were analyzed and, when necessary, described using absolute and relative frequencies. To calculate the percentage of agreement between reviewers, the Agreement Index was adopted, considering an acceptable agreement rate of 90% (14) . An exact test of binomial distribution was performed, indicated for small samples, with a significance level of p> 0.05 and a proportion of 0.90 of agreement, to estimate the statistical reliability of HF. Discrepancies in relation to the score were resolved through critical discussion among the reviewers.

RESULTS
The search resulted in 328 APPs, 250 in the Play Store and 78 in the App Store. After analyzing their titles, description of content and excluding duplicates, 316 were screened. After excluding those not related to the theme, 59 were selected for download and installed for complete evaluation. The APP selection process was described in Figure 1.
The themes (descriptions/objectives) were diverse and included knowledge (educational), self-care (management, selfmanagement, monitoring and recognition of signs and symptoms and recording of physical activities), decision-making (choice of treatment and assistance guides), follow-up of patients (professional follow-up at the outpatient clinic and follow-up by the multidisciplinary team after hospital discharge), calculators for patients with HF (diagnosis, complications, prognosis and mortality/hospitalization) and Systematization of Nursing Care (care plans) ( Table 1) .
APPs were targeted at health professionals, students and patients and their family members/caregivers. Among health professionals, two (5.3%) were specific to primary care physicians and four (10.6%) to nurses. Three (7.9%) APPs were also directed to university students in the health field and only two (5.3%) included family members/caregivers, although many patients with HF need their support to carry out their daily care.
The theme of knowledge was explored in 13 (34.2%) APPs; the themes of self-care and decision-making were found in seven (18.4%) each; six (15.8%) APPs were developed to prepare calculations for patients with HF; three (7.9%), for patient follow-up; two (5.3%) brought the Systematization of Nursing Care. CardioVisual and CardioExpert I had more than 100,000 downloads and were the most evaluated. CardioVisual is aimed at healthcare professionals and patients and is in version 5.1.6; has 3.6 stars out of 869 reviews. CardioExpert I is version 7.6.235, with 4.9 stars out of 1,129 reviews. Other APPs with more advanced versions were Heart Failure Manager (v.10.5.1), Heart Failure Storylines (7.17) and HFPath (v.6.1). Table 2 discusses information on the content, features, and usability of the selected APPs. The contents covered involved, mainly, habits (physical activity, weight control, feeding and water intake), well-being (living with the disease, mood and cognitive assessment), clinical data (anatomy, pathophysiology of the disease, vital signs and treatment) and disease management (management, postdischarge, multiprofessional follow-up and financial planning).
APPs contained various features: general, data insertion, diagnostics and those that require internet connection to control the management panel. Among the general ones, the use of texts (44.7%), figures (21.1%) and videos (18.4%) was more common. In the data insertion features, users' profile data (15.8%) and clinical data (55.3%). As for the resources that required internet connection, the use of links (4; 10.6%) and sending e-mails to health professionals (4; 10.6%) were more observed.
In the usability analysis by SUS, it was observed that 34 APP (89.5%) scored with a score > 68 points, instrument cutoff score, and 16 (42.1%) score above 85. The agreement rate among the observers was 92.1%, with p>0.05 for all items of SUS. Four (10.5%) APP obtained a borderline score (between 50 and 67 points), but still acceptable. The APP that did not reach the cutoff point were Patient Education Atlas of Heart Failure (50 points), Heart Failure Manager (62.5 points), Systolic Heart Failure (62.5 points) and Clinical Cardiology (52.5 points). Scores below 50 are considered without usability.
Based on SURE, the agreement rate among the observers was 94.7%, with p>0.05. It was found that four (10.5%) APP were at level 30 (Patient Education Atlas of Heart Failure, Clinical Cardiology, Cardiac Care Plans and Cardiac Nursing Care Plans); two (5.3%) at level 40 (Systolic Heart Failure and Heart Failure Info); seven (18.4%) at level 50; eight (21.1%) at level 70; 17 (44,7%) at level 80.  of Benchmarking of mobile apps on heart failure Cestari VRF, Florêncio RS, Garces TS, Pessoa VLMP, Moreira TMM.

DISCUSSION
The growth of the Internet increases the sale of smartphones and, with this, studies of construction and validation of APP, which already exceeds 165,000 (15) , portraying interest in the development of mobile technologies, collaborating in the construction of a new modality of health care (4) . This statement becomes clearer when observing the increasing construction of APP over the years.
Mobile APPs are shown to be innovative health care technologies. They are educational resources that allow improvement of teaching-learning, applied in different contexts (16) . In this sense, APPs, focusing on promoting patient knowledge, aims to meet the implementation of self-care practices (17) . In HF, self-care is vital to successful management. In an almost experimental research, HFApp, app for monitoring symptoms, reminders, education and screening of physiological data, proved effective to self-care, but was not significant in terms of awareness of the symptoms of the disease (6) .
The APP enables adequate follow-up of patients and assists in the clinical decision-making process of professionals, contributing to the development of reliable diagnoses and targeted therapeutic guidelines/conducts (3) , in addition to remote consultations (18) . In order to test the viability of a teleguidance APP with 692 german HF patients, a prospective study showed the technology as promising because it continuously reflected patient health information daily (19) . Thus, it allowed effective monitoring by health professionals, in addition to guiding decision-making.
With this technology, we can see the use by all those involved in the care process, in order to promote comprehensive care. When used by health professionals, it enables evidence-based practice. Therefore, it is relevant that professionals appropriate these tools to strengthen care and invest in safer care (17) . When developed for patients, APPs work as strategies that facilitate self-care, maintain autonomy and independence. For family members/caregivers, who play an important role in the care of HF patients, APPs are an additional form of information (5) .
Family members of patients with chronic diseases, such as HF, use the Internet to seek medical information more than the general public; on the other hand, use app (20) less. One justification found by these authors is that, possibly, family caregivers do not have enough time or find it difficult to use mobile APP focused on care. Sociodemographic factors and poor health and digital literacy are also associated with limited access to and use of APP in health. However, when they use them, they become more informed and empowered (21) .
The development of an APP involves more than producing and delivering the product ready for the customer, requires a whole procedure that covers, including, the characterization of versions. Semantic versioning (SemVer) is a set of particularity rules that show how app version numbers were developed, that is, the amount of changes you have made and which changes were compatible or not with the previous version (22) . It is a set of good practices and customs of software development that indicates the larger version (major), minor version (minor) and patch version, arranged among ready: major.minot.patch.
SemVer is important to give the customer a notion of the software's stability status, enabling you to identify whether the new version has new features or bug fixes (23) . Thus, users can know what to expect when they will update the software (24) . In the APP analysed, increasing versions were observed, indicating adaptation of contents with terms and words used worldwide, such as the addition of functionalities. Thus, it is perceived that SemVer is a break from the way of being moved by keeping the APP up to date.
There was a number of varied downloads, a value calculated by the number of users who enter the APP page and download. However, this number does not mean number of mobile phones with the APP installed. Some factors can influence the count of downloads, such as competition between search engines. The best way to minimize this problem is to invest in both stores to host the APP. In addition, having both versions is recommended for good performance, in addition to the reach of a larger number of people without restriction of use on only one of the operating systems.
In addition, APPs presented a variety of functions, essential for disease management. Although HF has no cure, changes in lifestyle can increase patients' quality and life expectancy. Continuous monitoring and care of daily vital signs allows to recognize changes or complications early. However, self-monitoring rates are low, as patients often forget to record relevant information, such as signs and symptoms, vital signs, and medications in use (4) . This complicates check-ups and makes it difficult to identify your condition's worsening.
Thus, APP that provide resources that offer a comfortable use for patients, with interactive elements that provide reliable information, such as data insertion, can increase patient safety, decrease hospitalization episodes, and share their data with health professionals. APP with risk stratification and diagnostics are essential tools for health professionals and patients for providing prognosis and personalized treatment (25) .
When designing an interactive system, it is necessary to seek an understanding of users' needs, so that one can meet, with objectivity and quality, the desired experience in performing the task that will be projected (26) . The criterion regarding the material relevance and its applicability is important if it presents a valid and understandable content for a target audience, but does not have viable and relevant applicability, this material needs to be critically rethought (27) .
Usability is a prerequisite for spreading mobile APP use to health and is defined as a set of software attributes that are based on the effort required for the use and individual evaluation of such use by an implicit set of users. The usability score is intrinsically correlates with better reliability of (28) content. Added to these factors are the lack of confidence in technology, frustration with design and navigation features, and an interest in having technology to support your self-management of the (29)(30) disease.
The use of mobile APP usability questionnaires is important because of special components such as connectivity issues, battery and security and privacy challenges (31) . There are several ways to conduct a usability study, being the use of questionnaires, the fastest and most practical way. The usability of the APPs analyzed was evaluated by two independent researchers using two instruments: SUS, which evaluates the general context (11) and SURE, smartphone-specific (13) . According to the SUS instrument, the vast majority of APPs scored above the cut-off score of the instrument.
Although SUS is widely used, this instrument was not designed to evaluate the usability of mobile health APP (31) . Thus, it was also decided to use SURE. It was found that the majority were at level 80 of sure, level at which respondents begin to fully agree that the help /tip given by the APP was useful, in addition to anure with the other items. This demonstrates the high level of satisfaction, which establishes intelligibility, learning, operability, attractiveness, and compliance with the (32) usability goals.
The repercussion of APPs for HF is vast and relevant, as it includes important themes, directed to all those involved in the care process. Therefore, it is necessary to support development and improvement, with a view to improving health practices. However, it was observed that themes such as vulnerability in health, mental health, sexual activity, social and family support, palliative care, side effects of medications, vaccinations and family planning were not addressed.
It was observed that, among health professionals, only physicians and nurses were contemplated. Multidisciplinary team support plays a crucial role to improve the quality of life of patients and family members since the diagnosis of the disease (33) , being the gold standard for monitoring patients with HF and its multiple comorbidities. This includes physicians, nurses, psychologists, nutritionists, pharmacists, dentists, physiotherapists, physical educators and social workers (34) . Furthermore, it is important to highlight the absence of a caregiver as a user.

Study limitations
It is worth noting that the research presented limitations regarding searches, such as the non-inclusion of paid APP and/ or that required registration by the institution or health professional, making it impossible to handle.

Contributions to nursing and health
Furthermore, the results strengthen the current knowledge about the APP available for download and handling about HF, by exploring the thematic scope and usability. Furthermore, it makes possible the identification of gaps, aiming at the development of new APPs that address relevant and indispensable themes, together with the appropriate usability. Moreover, it directs the use of resources that assist professionals, patients and family/caregivers in the involvement and adhering of care and in clinical decision-making.

FINAL CONSIDERATIONS
The APPs currently available on HF have comprehensive content and adequate usability that can guide patients, family members/ caregivers, health students, physicians and nurses. However, the findings showed the need to develop PAAs with more themes, such as vulnerability in health, palliative care, mental health, social support, sexual activity, essential in the care of patients with HF, in addition to the integration of other professionals from the multidisciplinary team.